Hotel Lobby AI Character Swap: Cast Your Duo
Two photos, two performers, one orange booth. Put yourself and a friend into the Hotel Lobby format, with a clear role for each photo.
How does Hotel Lobby AI character swap work?
Upload one photo for each of the two performers. The generator uses those images to create a Hotel Lobby AI remake guided by the preset performance. You choose the cast and output quality; the scene and motion reference are already supplied.
Make a two-character swap in four steps
1. Choose your duo
For a friends’ duet, use your own portrait in Photo 1 and your friend’s in Photo 2. You take the opening left-hand role; your friend takes the right-hand role. This is a casting example, not a generated result.
2. Check each photo
Keep one subject in each image, with a visible face and the outfit you want to reference. Inspect both upload previews and switch the photos if you want to reverse the roles.
3. Choose quality and generate
Select 480p, 720p or 1080p. Check the displayed credit cost and your balance, then submit from the generator.
4. Review both characters
Check the opening, turns and ending for identity changes or merged faces. Download the completed clip after reviewing the result.
Use the Hotel Lobby AI video generator to try this workflow. Read the preview and credit guide before spending credits.
Reference preview: the orange-booth format
Reference sample only — this is not generated from your uploaded photos.
Your two photos, assigned to two roles
Photo 1
Performer starting on the viewer’s left
Photo 2
Performer starting on the viewer’s right
The role follows the original performer through the clip. It is not reassigned whenever someone moves across the frame.
Start with the pair, then choose the photos
Pick a duo that makes sense to your audience: you and a friend, two colleagues, or two original characters. For example, put the friend you want in the opening left-hand role in Photo 1 and the other in Photo 2. Decide the casting before uploading so you can tell immediately if the result has assigned the roles correctly.
A useful reference shows more than a face
A clear portrait gives the model facial information. A waist-up or full-body photo also shows the outfit and silhouette you want it to use. Choose a single subject with a visible face; crop away other people. If clothing matters, avoid a tiny face crop that gives no clue about the rest of the person. These are reference-selection guidelines, not a guarantee of exact clothing reproduction.
Keep photo order tied to the original performer
Photo 1 is assigned to the performer who starts on the viewer’s left; Photo 2 to the performer who starts on the right. The generation instructions keep each identity with that performer through turns and occlusions. If the performers move, their identities should not change just because their current screen positions change. To reverse the casting, exchange the two photos before submitting.
Make the swap in the existing generator
Open the generator, add both photos, and inspect the two previews. Select 480p, 720p or 1080p and review the displayed credit cost before pressing Generate. Sign in when prompted. The site supplies the motion reference and instructions; you do not need to upload the performance clip or paste a prompt. The current workflow requests a 13-second video.
Review the whole performance, not just the first frame
Watch the opening, a head turn or hand gesture, and the ending. Ask whether each face is still recognizable, whether the outfit belongs to the right person, and whether both performers remain separate. Original audio is added during finishing; keeping that recording does not guarantee perfect lip synchronization. If a result is technically complete but the likeness is weak, a different reference may help, but a new generation uses the displayed credits.
Ready to try your duo? Open the generator or compare plans and credit packs.